Comparison · 8 min read
Betting strategies compared: what the maths says about each one
Most “betting strategies” are staking plans or tips, and neither changes the price you bet at. A few approaches are built on prices instead. Here they are side by side on five criteria, with the maths behind each verdict.
Updated 22 October 2026
Contents
The five criteria
- Expected value (EV). The average result per euro staked over many bets. It depends only on the prices you get and the true probabilities, never on how you size or order your stakes.
- Variance. How far results swing around that average. High variance means long losing runs even with positive EV.
- Effort. Time per day and how much you have to know.
- Capital. How much money the method needs to work as intended.
- Limit risk. How likely bookmakers are to restrict your account once you use the method consistently.
Comparison table
| Approach | Expected value | Variance | Effort | Capital | Limit risk |
|---|---|---|---|---|---|
| Martingale and other progressions | negative (the margin) | extreme: rare, very large losses | low | very high | low, but maximum stakes stop the system |
| Flat stakes on your own picks | negative for most people | medium | medium | low | low unless you win |
| Paid tips and tipsters | usually negative after fees; hard to verify | medium to high | low | low to medium | medium if the tips are good |
| Value betting | positive if the reference price is sharp | high | medium to high | medium (at least 100 units) | high |
| Sure bets | positive, small (about 1–2% listed) | low per bet; voids and mistakes add risk | high: fast, two bets at once | medium, spread over several accounts | high |
| Matched betting (bonuses) | positive per offer, finite | low | medium | low to medium | high (bonus bans); needs an exchange |
As of October 2026. Our assessment based on the maths below and the sources listed; your own results depend on prices, bookmakers and discipline, and can be negative.
Progressions: Martingale, Fibonacci, d'Alembert
A progression changes your stake after a win or a loss. Martingale doubles after every loss; Fibonacci moves one step along 1, 1, 2, 3, 5, 8, 13…; d'Alembert adds one unit after a loss and removes one after a win. They feel like strategies because they produce many small wins.
They can't change EV, for a simple reason: every individual bet still has the same expected value per euro. At odds of 2.00 on an outcome with a true probability of 48%, each euro staked returns 0.48 × 2.00 = €0.96 on average, whatever the stake. Over a series, the expected result is the total staked times −4%. Progressions don't remove the loss; they reshape it into many small wins and occasional very large losses. Our Martingale guide shows that after eight losses in a row the next stake is 256 times the first, and how often that happens.
The belief that a win is “due” after a losing run is the gambler's fallacy. Independent events don't remember previous results.
Tips, systems and accumulators
Flat staking on your own picks is the sensible way to size stakes, but staking doesn't create an edge. Against a typical 5% margin, you need to win 52.6% of bets at odds of 1.90 just to break even; our guide on winning long term shows what that means over 1,000 bets.
Paid tips add a subscription fee to the margin. Before you pay, ask for a complete, timestamped record of every tip, with the odds actually available at the time and the closing odds. Tipster results that show only wins, only screenshots or only percentages are not verifiable. If the tips really beat the closing line, your account will be limited sooner, which caps what they are worth.
Accumulators and system bets multiply margins: five legs at 95% return each give about 77% for the whole bet.
Price-based methods
Value betting backs prices above a fair estimate, usually derived from sharp bookmakers' odds with the margin removed. It has a positive expected value if that estimate is good, but one-sided bets lose often: at odds around 2.00 with a 4% edge, the standard deviation over 1,000 bets is about 32 units. Our guide to value betting covers the details. In ArbVantage, value bets are coming soon.
Sure bets cover every outcome of a market at different bookmakers when the prices add up to less than 100%. If every leg is placed at the shown odds and stands, the payout is about the same whatever the result: a small margin, not a guarantee. The work is speed and accuracy, the main risks are price changes, voids and limits. Details in sure betting explained.
Matched betting uses bookmaker bonuses and covers them with a lay bet at a betting exchange. It depends on a steady supply of offers and on exchange access. In Austria and Slovakia the major exchanges don't accept residents, which makes classic matched betting impractical there; converting free bets with back bets at other bookmakers, covered in our free bets guide, is the usual alternative.
Which approach fits whom
- You bet for fun: flat, small stakes, a fixed monthly budget, no progressions and few accumulators. Expect to lose about the margin.
- You want a method with positive expected value and can spend time daily: sure bets or free-bet conversion, with records from day one and the expectation that accounts get limited. Our bankroll guide covers how much capital that takes.
- You can live with long losing runs: value betting, with at least 100 units and closing line value as your check.
- You are thinking about Martingale: don't. The maths doesn't work, and the losses it produces are the large, rare kind.
Frequently asked questions
Which betting strategy is the best?
No strategy turns a negative-value bet into a positive one. Only approaches built on prices, such as sure bets, value bets and free-bet conversion, have a positive expected value, and each comes with costs: limits, variance, capital and time. Which fits you depends on how much time you have and how much variance you can live with.
Do progression systems like Martingale work?
No. Every bet keeps the same expected value per euro staked, so a progression cannot change the long-term result. It turns many small wins into rare, very large losses, and maximum stakes or an empty account stop it exactly when it needs to keep going.
Which strategy has the least risk?
Per bet, sure bets and free-bet conversion have the lowest variance, because every outcome is covered. They are not risk-free: prices can move before the second leg, bets can be voided and accounts get limited. Any method can still lose money.
Sources
- Wikipedia: Expected value
- Wikipedia: Martingale (betting system)
- Wikipedia: Gambler's fallacy
- Cain, M., Law, D., Peel, D. (2000): The favourite-longshot bias and market efficiency in UK football betting. Scottish Journal of Political Economy 47(1), 25–36
- Austrian Federal Ministry of Finance: gambling and player protection
As of October 2026. Ratings in the table are our assessment of the maths; all figures are worked examples.
Betting involves risk, including sure bets: prices move, bets get voided and accounts get limited. Only bet money you can do without. Our responsible gambling page lists free, confidential help.
Not betting, financial, legal or tax advice. 18+. As of October 2026.